If it wasn’t for his experience as a high school math educator, newly appointed Assistant Professor Luke Miratrix might never have focused on statistics.
“I was initially forced to teach statistics against my will when I was a high school teacher. In the course of teaching it, I fell in love with it in large part because it connected my interest in the real world with the abstract and beautiful world of mathematics,” he says. “It seemed a real tool to understand the world around me, kind of like a good painting or novel or philosophical work.”
Miratrix, formerly an assistant professor of statistics at Faculty of Arts and Sciences at Harvard University, has joined HGSE as an assistant professor.
“I’m really looking forward to working with a wide array of people on a variety of interesting and relevant problems,” he says. “I’m also really looking forward to helping people analyze data and helping people learn how to analyze data — it feels like a very direct way of having a positive impact on the scientific enterprise.”
Can you explain your research in layman’s terms?
My main research interests are coming up with simple methods for analyzing complex data. Simplicity is particularly relevant when trying to build causal arguments, which is obviously of great interest in most domains. I am also interested in how to use simulation to understand how good or reliable an estimate or conclusion based on data really is. I also really enjoy trying to extract information from text – here I attempt to pull fragments of text out of large corpera in order to hopefully shine light on some question of interest.
What drew you to this area of research?
Most of my research interests come from my being confused as to why and when the methods classically used to analyze data work. Even linear regression seems complicated to me, when I think about it deeply enough.
Why is it important for educators and practitioners to have a good understanding of statistics?
In these times it seems as if people who are comfortable with quantitative skills can hijack and railroad conversations about policy, science, or pretty much anything even when they are horribly, catastrophically wrong. Learning to not be intimidated and see what these arguments rest on is very important. On the flip side, because of the currently elevated status of quantitative argument, learning how to do it is important for defending one’s ideas. Learning how to engage on this level is a means to empowerment. Also, on a good day data analysis can actually help answer questions about how the world works. That is always a very satisfying feeling.